Stress effects on framed decisions: there are differences for gains and losses

Abstract

Recent studies have shown that acute stress can lead to riskier decision making. Yet, the underlying mechanisms of the stress effects on decisions under risk remain poorly understood. To gain a better understanding of decision-making processes and potential strategy application under stress, we investigated decision making in pure gain and loss domains with unequal expected values (EVs) across alternatives. We conducted an experimental study with a 2 × 2 design (stress vs. no stress and gain domain vs. loss domain). The Trier Social Stress Test (TSST) was utilized to induce acute stress. Controls performed the placebo-TSST (p-TSST). To validate the stress response we measured salivary cortisol and alpha-amylase concentrations. We used a modified version of the Game of Dice Task (GDT) to assess decision-making performance in a gain and a loss domain. Results showed that non-stressed participants made less risky decisions in the gain domain compared to those of the loss domain. This behavior is in accordance with previous studies and indicates the stability of the framing effect in even more complex tasks with changing EVs across alternatives. Stress did not alter risk taking behavior in the gain domain. Yet, in the loss domain stressed participants made less risky decisions compared to controls. Additionally, the data support earlier findings of longer reaction times in loss compared to gain domains due to higher cognitive effort for loss-framed decisions. It is discussed that stress may lead to reduced amygdala activation, which has been found to reduce riskier decisions in a loss domain. With respect to earlier results of riskier decisions in tasks that unite both gain and loss domains, it is discussed whether stress leads to a stronger evaluation of high gains and a neglect of losses.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 81 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.